Uterine leiomyomas (ULs) are extremely common, benign smooth muscle tumors. Their incidence is 70% in the Caucasian population by the age of fifty and even higher in black women. They cause symptoms such as pain and abnormal menstrual bleeding in 25% of the affected women. At present, surgery is still the only widely used curative treatment option. Taken together, ULs cause both significant morbidity and a huge economic burden. Based on driver alterations, ULs have been divided into three distinct subtypes: 70-80% of ULs have a MED12 mutation, 10-15% a HMGA2 overexpression, and 1% a biallelic FH inactivation. The driver event for the remaining 10-15% of ULs remains unknown. The subtypes have different molecular and clinical characteristics, but the tumorigenic mechanism of the driver events is elusive. Based on family studies, population disparity, and previous linkage- and association analyses, ULs have a hereditary component. The aim of this thesis was to characterize both the germline and somatic landscape of ULs utilizing various genome-wide methods and both biobank material and a large collection of over two thousand prospectively collected tumor samples. Previous studies on the heritable component of ULs have been limited by small sample sizes and have discovered only few significant loci. Furthermore, the role of these loci in myomagenesis remains unclear. In Study I, we performed the largest UL genome-wide association study (GWAS) to date revealing 22 significant loci. Most of the candidate genes could be categorized as either genitourinary development genes or genetic stability genes. Among the candidate genes were ESR1, MED12, TP53, and TERT. In Study II we scrutinized a set of tumors with an unknown driver event and discovered mutations in six of the nine SRCAP complex genes. Overall, these mutations were detected in 2% of all ULs. SRCAP complex loads variant histone H2A.Z into chromatin, mostly to active and bivalent transcription start sites and enhancers. We showed that this loading is defective in the SRCAP complex mutated lesions. By integrating data from various sources, we were also able to highlight the importance of bivalent regulation in all UL subtypes and pinpoint genes regulating development and differentiation, such as CBX2, CBX4, and CBX8, as potential players in UL genesis. In Study III we examined SNP-array data from 1,946 tumors and showed an association between parity and the total length of allelic imbalance, complex chromosomal rearrangements, and whole chromosome losses. Mechanical forces have been shown to be able to cause DNA damage; in pregnancy and parturition, the uterus undergoes major changes in size and is affected by large mechanical forces. Furthermore, we showed in vitro that stretching can cause damage and detected both increased DNA damage and repair in stretched myoma and myometrial cells. Taken together, we established mechanical stretching as a potential new mechanism for tumorigenesis. The differential diagnosis of uterine tumors is sometimes complicated. In Study IV we detected nine tumors initially diagnosed as ULs expressing tyrosine kinase fusion genes. Such fusions have not been described in ULs but are typical for inflammatory myofibroblastic tumors (IMTs). IMT is a rare tumor type with intermediate malignant potential. These tumors can have a fascicular growth pattern that closely resembles ULs. We discovered that the protein expression and morphological pattern of the tyrosine kinase fusion-positive samples were leiomyoma-like. In gene expression level, some differences to UL subtypes were detected. Among the pathways enriched in the tyrosine kinase fusion-positive lesions was epithelial-to-mesenchymal transformation. Our results raise the question on whether these tumors are UL-like IMTs or tyrosine kinase fusion-positive ULs and highlight the heterogeneity of uterine lesions.
|Tila||Julkaistu - 2022|
|OKM-julkaisutyyppi||G5 Tohtorinväitöskirja (artikkeli)|
LisätietojaM1 - 121 s. + liitteet
- 3111 Biolääketieteet
- 1184 Genetiikka, kehitysbiologia, fysiologia